RESEARCH AND PRACTICE
TABLE 2—Unstandardized Regression Coefficients (SE) Predicting Physical Health
Physical Health
Age Income Education (1 = bachelor’s degree) Marital status (1 = married) John Henryism Constant R2
–.096 (.021)† .000 (.000)† –1.14 (.682)* –1.32 (.760)* –.121 (1.00)* 49.17 (3.61)† .140
pretation of the findings. S. L. Sellers conducted the statistical analyses. H. W. Neighbors provided guidance in the interpretation of the findings. Each author contributed equally to critical revision of the article.
Swimmer’s Itch: Incidence and Risk Factors
Acknowledgments
| Lois M. Verbrugge, PhD, MPH, Jeanette J.
Primary support for this study was provided by a grant from the Centers of Excellence Small Research Grants Program, Health Resources and Services Administration (project no. D34 MB 04036-06). The authors would like to thank the men of Omega Psi Phi Fraternity Inc, Tenth District, that participated in Project UPLIFT for their commitment to improving the health of African American men.
*P < .1; **P < .05; ***P < .01; †P < .001.
Human Participant Protection Some studies indicate that John Henryism protects upper-SES African Americans from the negative health effects of race-based stressors, whereas others indicate the opposite.10–12 This study finds that among high-SES African American men, John Henryism is beneficial for health. The reasons for the positive health effects of John Henryism remain the subject of conjecture. John Henryism is an individual behavioral predisposition that can pay major dividends for career achievement and material gain.13 Clearly, the economic and educational success of these men provides them greater access to quality health care. Our data indicate that in the context of high SES, John Henryism is a resource that African American men draw on to contribute to their positive health outcomes. We speculate that John Henryism is conducive to increasing personal responsibility for one’s health with the same single-minded determination to succeed. This hypothesis deserves further study. To develop public health strategies to improve the health of all African American men requires the study of the social and behavioral implications of health of men of differing SES.
About the Authors At the time of the study Vence L. Bonham was with Michigan State University, East Lansing. Sherrill L. Sellers is with the University of Wisconsin, Madison. Harold W. Neighbors is with the University of Michigan, Ann Arbor. Requests for reprints should be sent to Vence L. Bonham, National Human Genome Research Institute, National Institutes of Health, 31 Center Drive Bldg. 31, Room 4B09, Bethesda, MD 20892 (e-mail:
[email protected]). This brief was accepted June 15, 2003.
Contributors V. L. Bonham was the principal investigator of the project and conceived the study. V. L Bonham, S. L. Sellers, and H. W. Neighbors each contributed substantially to conceptualization and design of the study and to inter-
This study was approved by the Michigan State University committee on research involving human subjects.
References 1. Eberhart MS, Ingram DD, Makuc DM, et al. Urban and Rural Health Chartbook. Health, United States, 2001. Hyattsville, Md: National Center for Health Statistics; 2001. 2. Williams DR, Neighbors HW. Racism, discrimination and hypertension: evidence and needed research. Ethnicity Dis. 2001;11:800–816. 3. James SA, LaCroix AZ, Kelinbaum DG, Strogatz DS. John Henryism and blood pressure differences among black men. II The role of occupational stressors. J Behav Med. 1984;7:259–275. 4. James SA, Hartnett SA, Kalsbeek WD. John Henryism and blood pressure differences among black men. J Behav Med. 1983;6:259–279. 5. James SA, Strogatz DS, Wong SB, Ramsey DL. Socioeconomic status, John Henryism and hypertension in blacks and whites. Am J Epidemiol. 1987;126:664–673. 6. James SA, Keenan NL, Strogatz DS, Browning SR, Garrett JM. Socioeconomic status, John Henryism, and blood pressure in Black adults: The Pitt County Study. Am J Epidemiol. 1992;135:59–67. 7. James, SA, Thomas, PE. John Henryism and blood pressure in Black populations: a review of the evidence. African Am Res Perspectives. 2000;6(3):1–10. 8. Duijkers TJ, DrijverM, Kromhout D, James SA. “John Henryism” and blood pressure in a Dutch population. Psychosom Med. 1988;50(4):353–9. 9. Ware JE, Kosinski M, and Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Medical Care. 1996;34(3):220–233. 10. Markovic N, Bunker CH, Ukoli FA, Kuller LH, John Henryism and blood pressure among Nigerian civil servants, J Epidemiol Community Health. 1998;52:186–190. 11. Light KC, Brownley KA, Turner JR, Hinderliter AL, et al. Job status and high-effort coping influence work blood pressure in women and blacks. Hypertension 1995;25:554–559. 12. McKetney EC, Ragland DR. John Henryism, education and blood pressure in young adults: The CARDIA study. Am J Epidemiol Community Health. 1998;52(3): 186–190. 13. Sellers SL, Neighbors HW. Goal striving stress, social economic status, and the mental health of black Americans. Ann New York Acad Sci. 1999;896:469–473.
738 | Research and Practice | Peer Reviewed | Verbrugge et al.
Rainey, MPH, Ronald L. Reimink, MA, and Harvey D. Blankespoor, PhD
Swimmer’s itch (cercarial dermatitis) affects people engaged in open-water activities. We report incidence and risk factors for a US lake. Water exposures and swimmer’s itch experience were reported daily for riparian household residents and guests at Douglas Lake, Michigan, in July 2000. Incidence of swimmer’s itch was 6.8 episodes per 100 water exposure days. Positive risks were (1) exposures in shallow water and in areas with onshore winds and (2) more days of lake use in July. Further epidemiological studies will help public health agencies address this bothersome problem at recreational lakes. (Am J Public Health. 2004;94:738–741)
Cercarial dermatitis, or swimmer’s itch, is a skin condition that affects people engaged in open-water activities in fresh and salt-water areas around the world. It causes intense discomfort, discourages recreational water use, and leads to economic loss for lake regions if people decide to vacation elsewhere. Since discovery of the parasites responsible for swimmer’s itch,1 most research has focused on the schistosomes’ life cycle and biological control methods.2–14 The literature on humans and swimmer’s itch consists largely of clinical and outbreak reports.15–28 Epidemiological studies are rare.29,30 We conducted a prospective study at a Michigan lake to obtain an incidence rate and identify risk factors for swimmer’s itch. The schistosomes for swimmer’s itch have a 2-host life cycle, an avian definitive host and a snail intermediate host. At one point in the cycle, snails release cercariae (freeswimming larval stage of the parasite) into the water, where they may encounter and enter ducks and other birds. If instead they penetrate human skin during recreation or work in
American Journal of Public Health | May 2004, Vol 94, No. 5
RESEARCH AND PRACTICE
the water, an inflammatory response occurs. Mild itching and macular eruptions occur 1 to several hours after a person leaves the water. Intense itching and papules are present 10 to 15 hours later, continuing for about a week. Swimmer’s itch is not communicable. Preventive actions before or after water exposure have been proposed, but no scientific evidence exists on their efficacy.
comfort. When a resident or guest departed, no data for subsequent days were entered. We developed an algorithm to detect episodes of swimmer’s itch: each day with new itchy spots was linked to water exposure characteristics the same or prior day. Preference was given to exposure that same day, due to the typically short incubation period for symptoms. New spots without same-day or prior-day water exposure were dropped.
METHODS RESULTS A prospective survey of swimmer’s itch among persons with water-based activities was conducted in July 2000 at Douglas Lake, Cheboygan County, Michigan. The lake covers 15.33 sq km and has a maximum depth of 24 to 25 m. Riparian (lake perimeter) households were the target sample. Inclusion criteria were that residents planned to be at the lake for 2 weeks or longer in July and used the lake regularly for swimming, wading, or other activity involving direct contact with the water. Project advertisements were posted around the lake, and households were recruited by personal visit in late June. Participating households filled in a diary for the month of July 2000, a peak period of water use and swimmer’s itch complaints. Water exposures and trouble with swimmer’s itch were recorded for all residents and guests at the household. The diary booklet had a page for each day, plus information sheets (e.g., swimmer’s itch criteria, lake map with numbered areas). Each day, residents and guests were listed by initials, age, and sex. Water exposure information was as follows: total minutes in water (swimming, wading, other recreation or work directly in Douglas Lake), minutes in shallow water (waist deep or less), lake area where most exposure occurred, time of day when most exposure occurred, and whether preventive action was taken (e.g., oil/lotion/wet suit before exposure, brisk brush/alcohol rub after). These items tap hypothesized risk factors for swimmer’s itch, namely, long duration in water, shallow-water exposure, locations with shallow shorelines or onshore winds, early- or mid-morning exposure, and human practices to avert or remove cercariae. Swimmer’s itch information was as follows: number of new itchy spots that day due to swimmer’s itch, total itchy spots that day, and degree of dis-
ates showed that location and preventive action remained significant.
Swimmer’s Itch Risk Factors: Exposed Persons Significant factors for incidence of swimmer’s itch for people who used the lake in July (n = 301) are shown in Table 1. Key bivariate risks were number of days of water exposure in July, minutes in water, and minutes in shallow water. Age and sex had no effect. Multivariate analyses confirm the importance of how many days people used the lake.
Descriptive Statistics Forty households completed the diary for July. The total number of residents and guests at the households was 313. Almost all residents and guests (301; 96.2%) were exposed to water in Douglas Lake in July. The 301 lake users reported 1300 water exposure days. On exposure days, mean time in the water was 49.0 minutes. Shallow-water use was common (89.2% of the days; mean = 39.8 minutes). Exposures occurred at 36 of the 54 lake areas; most (63.4%) were in the afternoon from 2 to 6 PM. Prevention was used on 29.5% of the days. The swimmer’s itch incidence rate was 6.8 episodes per 100 water-exposure days (89 episodes; 95% confidence interval = 5.5, 8.2). Fifty-two people acquired swimmer’s itch (17.3% of exposed persons); 58% had 1 episode and 42% had 2 or more.
Swimmer’s Itch Risk Factors: Water-Exposure Days Significant factors for incidence of swimmer’s itch for water-exposure days (n = 1300) are shown in Table 1. Key bivariate risks were any shallow-water use, location, and preventive action. All episodes occurred on days with shallow-water use. Incidence was highest in south and east lake zones (the 54 lake areas were pooled into 5 zones for analysis; see Table 1, footnote c). Paradoxically, using prevention was positively associated with onset of swimmer’s itch. Although observed incidence rose with total minutes and shallow minutes, and was highest for exposures from 6 to 10 AM, those results were not statistically strong (P > .10). Age and sex were unrelated to onset of an episode. Logistic regressions with all covari-
May 2004, Vol 94, No. 5 | American Journal of Public Health
DISCUSSION This study found the following risk factors for onset of swimmer’s itch: amount of water exposure, especially shallow-water use; exposure in several lake zones; and taking preventive action. We interpret these results in biologically relevant ways, as follows. (1) Shallow water is where snail beds are typically most dense and where cercariae tend to accumulate, so incidence is higher for people using shallow water. More days of lake use increase a person’s chances of encountering cercariae at any depth. (2) To explain location effects, we assembled existing knowledge about snail densities, duck broods, and weather patterns for Douglas Lake. The most likely reason for elevated incidence in the southern and eastern zones is that persistent winds brought cercariae from the northwestern and western areas, and the sheltered bays prevented the onward movement of incoming and locally produced cercariae. Many factors affect cercarial concentrations from year to year and place to place; our explanation for location differences in 2000 is possible but unproven. (3) Taking preventive action before or after exposure was linked to increased risk of swimmer’s itch. This aligns with evidence that people’s sensitivity to swimmer’s itch increases over time.14 Past bad experience may prompt people to use prevention, but it is apparently insufficient to avert new episodes. The low incidence rate of 6.8 episodes per 100 water-exposure days is probably related to weather conditions. Summer 2000 was uncommonly cool in northern Lower Michigan. A warmer summer would probably generate higher incidence because cer-
Verbrugge et al. | Peer Reviewed | Research and Practice | 739
RESEARCH AND PRACTICE
ties and worked closely with Verbrugge in interpretation and presentation of results. R. L. Reimink managed the project fieldwork, including household recruitment and contacts. H. D. Blankespoor assisted in project design and selection of Douglas Lake as the site. With longtime knowledge of the lake, Reimink and Blankespoor helped interpret location effects.
TABLE 1—Predictors of Swimmer’s Itch Incidence on Water-Exposure Days and for Exposed Persons: Swimmer’s Itch Project, Douglas Lake, Cheboygan County, Michigan, July 2000 Predictor
% of Water-Exposure Days With Onset or Correlation (r) With Any Episode in July
Swimmer’s Itch Onset on Days With Water Exposure Minutes spent in water NS Minutes spent in shallow water NS Any shallow water exposureb Yes = 7.6% vs No = 0.0%, † Location of water exposure (on lake)c South = 9.3% vs West = 4.4%, ** South = 9.3% vs Northwest = 2.6%, *** North = 8.0% vs West = 4.4%, * North = 8.0% vs Northwest = 2.6%, ** East = 7.5% vs Northwest = 2.6%, * Time of day NS Preventive action taken Yes = 9.4% vs No = 4.6%, *** Age NS Sex NS Any Episode of Swimmer’s Itch in July No. of days of water exposure in July .429, † Minutes spent in water during July .279, † Minutes spent in shallow water during July .394, † Age NS Sex NS
Odds Ratio From Logistic Regressiona
NS NS ... South vs West, 2.36, *** South vs Northwest, 3.89, *** North vs West, 2.05, * North vs Northwest, 3.38, ** East vs Northwest, 3.39, * NS Yes vs No, 2.22, *** NS NS
Acknowledgments This study was a pilot project to evaluate data collection methods and obtain preliminary results. The Douglas Lake Association approved the project and helped advertise it to households. Hope College, Holland, Mich, provided some funds for data analysis. The project team thanks Dr James M. Lepkowski (Institute for Social Research and Department of Biostatistics, University of Michigan) for consultation on sample size estimation for the project; Jim Laarman, Cliff Graves, and Harry Blecker of the Crystal Lake Association, Benzie County, Michigan, for discussions of swimmer’s itch over several years preceding the project; Aryc W. Mosher for initial design work; and Elmer G. Gilbert for design and analysis advice throughout the project.
1.21, † NS NS NS NS
Human Participant Protection This nonfunded pilot project followed contemporary survey practice for household recruitment, retention, and contact. Households could cease diary-keeping at any time (2 of 42 did so). No personal identifiers appear in the data.
Note. Number of water-exposure days = 1300; number of people with water exposure in July = 301. NS = nonsignificant (P > .10). a Minutes in water and minutes in shallow water are strongly correlated (r = 0.42 for water-exposure days; r = 0.74 for exposed persons), so regressions are computed with one, then with the other. Results for minutes in water are shown here. b Multivariate analyses are not possible because all episodes occurred on shallow-water exposure days. c Respondents recorded location of their water exposure using a map dividing the lake perimeter into 54 areas. For analyses, the areas were pooled into 5 zones (west, northwest, north, east, south) separated at shore points with bays in between. We tested all pairs of zones for significant differences; significant contrasts are shown here; all others are nonsignificant (NS). *P < .10; **P < .05; ***P < .01; †P < .001.
carial production and concentrations increase in warm conditions. For more information about methods and results, contact the lead author. A longer article including complex variance estimations is forthcoming.31
CONCLUSIONS The onset of swimmer’s itch depends on how humans interact with the lake. Exposures to shallow water and areas with onshore winds are key risks for swimmer’s itch. The more days a person used the lake in July, the higher his or her chances of having any episode. The underlying biological reasons relate to density of snails, movement of cercariae, and cumulative chances of encountering cercariae. In public health
References 1. Cort WW. Schistosome dermatitis in the United States (Michigan). JAMA. 1928;90:1027–1029. 2. Blankespoor HD, Reimink RL. The control of swimmer’s itch in Michigan: past, present, and future. Mich Academician. 1991;24:7–23. 3. Leighton BJ, Zervos S, Webster JM. Ecological factors in schistosome transmission, and an environmentally benign method for controlling snails in a recreational lake with a record of schistosome dermatitis. Parasitol Int. 2000;49:9–17.
terms, people should avoid using shallow water and areas with persistent onshore winds. If they still choose such places, the less often, the better.
4. Muller V, Kimmig P, Frank W. The effect of praziquantel on Trichobilharzia (Digenea, Schistosomatidae), a cause of swimmer’s dermatitis in humans [in German]. Appl Parasitol. 1993;34:187–201.
About the Authors Lois M. Verbrugge is with the Institute of Gerontology and Jeanette J. Rainey is with the Department of Epidemiology, University of Michigan, Ann Arbor. Ronald L. Reimink is with the Science Department, Hudsonville High School, Hudsonville, Mich. Harvey D. Blankespoor is with the Department of Biology, Hope College, Holland, Mich. Requests for reprints should be sent to Lois M. Verbrugge, PhD, MPH, Institute of Gerontology, 300 N Ingalls, University of Michigan, Ann Arbor, Michigan 48109–2007 (e-mail:
[email protected]). This brief was accepted May 4, 2003.
5. Reimink RL, DeGoede JA, Blankespoor HD. Efficacy of praziquantel in natural populations of mallards infected with avian schistosomes. J Parasitol. 1995;81: 1027–1029. 6. Blankespoor HD. Patterns of swimmer’s itch in the northern part of the Lower Peninsula of Michigan, USA. J Med Appl Malacol. 2000;10:47–55. 7. Keas B, Blankespoor HD. The prevalence of cercariae from Stagnicola emarginata (Lymnaeidae) over 50 years in northern Michigan. J Parasitol. 1997;83: 536–540.
Contributors
8. Cort WW. Studies on schistosome dermatitis, XI: status of knowledge after more than twenty years. Am J Hyg. 1950;52:251–307.
L. M. Verbrugge was principal investigator, with chief responsibility for project design and analysis. J. J. Rainey performed all data management and computing activi-
9. Burton TM, Muzzall PM, Snider RJ, Coady NR. Swimmer’s itch in Michigan: an outlook from Michigan State University. Mich Riparian. November 1998:8–9.
740 | Research and Practice | Peer Reviewed | Verbrugge et al.
American Journal of Public Health | May 2004, Vol 94, No. 5
RESEARCH AND PRACTICE
10. Muzzall PM, Burton TM, Snider RJ, Coady NR, Saxton J, Sergeant M. Swimmer’s itch in Michigan: another outlook from Michigan State University, part II. Mich Riparian. August 2000:13–15. 11. Loken BR, Spencer CN, Granath WO Jr. Prevalence and transmission of cercariae causing schistosome dermatitis in Flathead Lake, Montana. J Parasitol. 1995;81:646–649. 12. Scott ME, Burt MD. Swimmers’ itch in New Brunswick: distribution and description of the causative agent, Cercaria catascopii n.sp. Can J Zool. 1976;54: 2200–2207.
30. Lindblade KA. The epidemiology of cercarial dermatitis and its association with limnological characteristics of a northern Michigan lake. J Parasitol. 1998;84: 19–23. 31. Verbrugge L, Rainey JJ, Reimink RL, Blankespoor HD. Prospective study of swimmer’s itch incidence and severity. J Parasitol. In press.
METHODS
13. Kimmig P, Meier M. Parasitologic studies, diagnosis and clinical aspects of cercarial dermatitis: public health significance for bathing waters in temperate zones [in German]. Zentralbl Bakteriolog Mikrobiolog Hyg [B]. 1985;181:390–408.
Cancer Burden From Arsenic in Drinking Water in Bangladesh
14. Olivier L. Schistosome dermatitis, a sensitization phenomenon. Am J Hyg. 1949;49:290–302.
| Yu Chen, MPH, and Habibul Ahsan,
15. Baird JK, Wear DJ. Cercarial dermatitis: the swimmer’s itch. Clin Dermatol. 1987;5:88–91.
MD, MMedSc
16. Mulvihill CA, Burnett JW. Swimmer’s itch: a cercarial dermatitis. Cutis. 1990;46:211–213.
We assessed the potential burden of internal cancers due to arsenic exposure in Bangladesh. We estimated excess lifetime risks of death from liver, bladder, and lung cancers using an exposure distribution, death probabilities, and cancer mortality rates from Bangladesh and dose-specific relative risk estimates from Taiwan. Results indicated at least a doubling of lifetime mortality risk from liver, bladder, and lung cancers (229.6 vs 103.5 per 100 000 population) in Bangladesh owing to arsenic in drinking water. (Am J Public Health. 2004;94:741–744)
17. Appleton CC, Lethbridge RC. Schistosome dermatitis in the Swan estuary, Western Australia. Med J Aust. 1979;1:141–144. 18. Gonzalez E. Schistosomiasis, cercarial dermatitis, and marine dermatitis. Dermatol Clin. 1989;7: 291–300. 19. Hoeffler DF. “Swimmers’ itch” (cercarial dermatitis). Cutis. 1977;19:461–465, 467. 20. Kirschenbaum MB. Swimmer’s itch. A review and case report. Cutis. 1979;23:212–217. 21. Centers for Disease Control. Cercarial dermatitis among bathers in California [and] Katayama syndrome among travelers to Ethiopia. MMWR Morb Mortal Wkly Rep. 1982;31:435–438. 22. Centers for Disease Control. Cercarial dermatitis outbreak at a state park, Delaware, 1991. MMWR Morb Mortal Wkly Rep. 1992;41:225–228. 23. de Gentile L, Picot H, Bourdeau P, et al. Cercarial dermatitis in Europe: a new public health problem? [in French] Bull World Health Organ. 1996;74:159–163. 24. Effelsberg W. Duck bilharziasis in the medical anthropologic perspective [in German]. Offentl Gesundheitswes. 1989;51:123–127. 25. Knight R, Worms MJ. An outbreak of cercarial dermatitis in Britain. Trans Royal Soc Trop Med Hyg. 1972;66:21. 26. Kolarova L, Skirnisson K, Horak P. Schistosome cercariae as the causative agent of swimmer’s itch in Iceland. J Helminthol. 1999;73:215–220. 27. Kullavanijaya P, Wongwaisayawan H. Outbreak of cercarial dermatitis in Thailand. Int J Dermatol. 1993; 32:113–115. 28. Levy DA, Bens MS, Craun GF, Calderon RL, Herwaldt BL. Surveillance for waterborne-disease outbreaks, United States, 1995–1996. MMWR Morb Mortal Wkly Rep. 1998;47(SS-5):1–34. 29. Chamot E, Toscani L, Rougemont A. Public health importance and risk factors for cercarial dermatitis associated with swimming in Lake Leman at Geneva, Switzerland. Epidemiol Infect. 1998;120:305–314.
tion of Bangladesh.1 The goal of the present study was to estimate excess lifetime mortality rates for the most-established arsenicrelated internal cancers (i.e., lung, liver, and bladder cancers)7–10 in Bangladesh.
Groundwater contamination caused by inorganic arsenic is a massive public health hazard in Bangladesh.1–4 The millions of handpumped tube wells installed since the 1970s have led to 95% of the country’s 130 million residents becoming dependent on supposedly pathogen-free underground water.5 It is estimated that 25 to 57 million people in Bangladesh have suffered chronic exposure to arsenic,1,5 and because decades of exposure have already accrued, the exposed population is at an elevated risk of arsenic-induced health problems. The principal cause of arsenic-induced mortality is cancer,6–10 but little is known regarding future cancer mortality risks attributable to arsenic exposure among the popula-
May 2004, Vol 94, No. 5 | American Journal of Public Health
Calculation of lifetime excess risks due to a particular exposure requires measures of distribution of the exposure, “background” lifetime risks, and dose-specific relative risk estimates. In the present study, these measures were estimated as follows. The arsenic exposure distribution in Bangladesh was ascertained from a sample of 65 876 people who represented the source population of an ongoing prospective cohort study focusing on the health effects of exposure to arsenic in drinking water. Water samples from 5966 contiguous hand-pumped tube wells in a well-defined geographic area of Araihazar, Bangladesh, were collected and tested for arsenic in 2000. Well owners were interviewed to collect data on the numbers and characteristics of the 65876 regular users.11 Gender-specific lifetime mortality risks from liver, bladder, and lung cancers among the population of Bangladesh were derived, via life table methods, from the formula ∑S(tk )Pk. Values of S(tk ) indicate the probability of surviving to the beginning of each of the 5 (i.e., k = 1–5) age groups assessed (0–14, 15–44, 45–54, 55–64, ≥65 years). Survival estimates were based on genderand age-specific death probabilities among the overall population of Bangladesh.12 Values of Pk indicate gender-, age-, and cancer-specific mortality rates in Bangladesh; these rates were computed by the International Agency for Research on Cancer (IARC).13–16 Gender-specific, age-adjusted relative risks of liver, bladder, and lung cancer mortality due to arsenic exposure were computed on the basis of gender- and age-specific data on arsenic exposure, cancer mortality, and at-risk population obtained from studies conducted in Taiwan (detailed data regarding a published study17 were obtained from C. J. Chen and L. Ryan, January 2002). We used Poisson regression models in calculating these risk estimates, allowing us to compare rates for different levels of arsenic exposure in an en-
Chen and Ahsan | Peer Reviewed | Research and Practice | 741
